WO1996016380A1 - Systeme et procede d'interpolation adaptative de donnees video - Google Patents

Systeme et procede d'interpolation adaptative de donnees video Download PDF

Info

Publication number
WO1996016380A1
WO1996016380A1 PCT/US1995/013560 US9513560W WO9616380A1 WO 1996016380 A1 WO1996016380 A1 WO 1996016380A1 US 9513560 W US9513560 W US 9513560W WO 9616380 A1 WO9616380 A1 WO 9616380A1
Authority
WO
WIPO (PCT)
Prior art keywords
inteφolation
kernel
image data
kernel lookup
image
Prior art date
Application number
PCT/US1995/013560
Other languages
English (en)
Inventor
Abolghassem B. Mahmoodi
Original Assignee
Minnesota Mining And Manufacturing Company
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Minnesota Mining And Manufacturing Company filed Critical Minnesota Mining And Manufacturing Company
Priority to JP8516854A priority Critical patent/JPH10509824A/ja
Priority to EP95938826A priority patent/EP0793836A1/fr
Priority to AU40062/95A priority patent/AU4006295A/en
Publication of WO1996016380A1 publication Critical patent/WO1996016380A1/fr

Links

Classifications

    • GPHYSICS
    • G06COMPUTING; CALCULATING OR COUNTING
    • G06TIMAGE DATA PROCESSING OR GENERATION, IN GENERAL
    • G06T3/00Geometric image transformation in the plane of the image
    • G06T3/40Scaling the whole image or part thereof
    • G06T3/4007Interpolation-based scaling, e.g. bilinear interpolation
    • G06T3/10

Definitions

  • the present invention relates to image processing techniques, and, more particularly, to interpolation techniques for expansion of digital image data for presentation on output media. Discussion of Related Art
  • a digital laser imager forms a visible representation of an image on imaging media, such as photosensitive film, based on a set of image data representative of the image.
  • the image data contains a plurality of digital image values acquired by an input imaging device.
  • the input imaging device may comprise, for example, a diagnostic device, such as a magnetic resonance (MR), computed tomography (CT), conventional radiography (X-ray), or ultrasound device.
  • MR magnetic resonance
  • CT computed tomography
  • X-ray conventional radiography
  • ultrasound device ultrasound device.
  • the image data may represent an image of a physiological object obtained for diagnostic purposes.
  • Each of the digital image values corresponds to one of a plurality of pixels in the original image, and represents an optical density associated with the respective pixel.
  • the laser imager processes the digital image values to generate laser drive values.
  • Each of the laser drive values represents an exposure level calculated to reproduce, on the imaging media, the optical density of a pixel in the original image.
  • the laser drive values are used to modulate the intensity of a scanning laser, thereby exposing imaging media with a particular level of exposure.
  • the laser imager subsequently develops the imaging media to form the visible representation of the original image.
  • the number of pixels represented by the image data can be significantly less than the resolution of the imaging media.
  • laser imagers include scanning lasers capable of producing an output resolution of 300 pixels per inch on the imaging media.
  • the original image may be represented by image data comprising as little as a 256 by 256 array of pixels.
  • the laser imager may provide different formats that enable a user to select the number of images to be formed.
  • the number of images formed on the imaging media determines the size of each image.
  • a 512 by 512 pixel set of image data must be expanded by at least twenty-five percent to match the 300 pixels per inch resolution of the laser imager.
  • formats requiring larger sizes it is necessary to expand the image data by much greater amounts.
  • formats producing smaller sizes require that the image data be reduced.
  • Expansion involves resampling of the originally sampled image data at a higher rate dependent on the amount of expansion required.
  • Reduction involves resampling of the originally sampled image data at a lower rate dependent on the amount of reduction required.
  • Resampling is carried out by performing interpolation on the original image data to effectively fit a curve to the original image data.
  • the curve is resampled at a higher rate than the original image data to create intermediate pixel points between those available from the originally sampled image data.
  • For reduction the curve is resampled at a lower rate than the original image data.
  • the resampling process transforms the image data into an image having a higher spatial resolution for expansion and a lower spatial resolution for reduction.
  • the resampling rates for expansion can degrade image quality by loss of high frequency information, introducing perceived blurring and other artifacts into the expanded image data.
  • the ideal low-pass filter would remove replicates of the frequency introduced by the resampling.
  • the implementation of an ideal low-pass filter is not feasible, however, because it would require the availability of many surrounding data points for each interpolated pixel.
  • interpolation functions seek to approximate the effect of an ideal low-pass filter.
  • Such interpolation functions generally fall into two categories, commonly referred to as (1) “two-point” interpolation and (2) “four- point” interpolation.
  • One type of "two-point” interpolation involves "pixel replication” interpolation, whereby sample points between two adjacent original pixels are simply assigned the intensity values of the closest adjacent original pixel.
  • Another type of "two-point” interpolation involves "bilinear” interpolation, whereby sample points between two adjacent original pixels are assigned linearly weighted intensity values based on distance from the closest adjacent original pixel.
  • Other "two-point” interpolation techniques are implemented by third order, fifth order, seventh order, etc., polynomial kernel functions.
  • “Four-point” interpolation may involve the application of a cubic spline function.
  • the four-point interpolation carried out by application of the cubic spline function involves much more computation than two-point interpolation.
  • the cubic spline function has been found to produce a better spatial frequency response than two-point interpolation functions in most cases, providing a closer approximation of the ideal low-pass filter spectrum.
  • a cubic spline interpolation function may produce a less than optimum interpolated image, relative to two-point interpolation functions such as bilinear interpolation or pixel replication.
  • United States Patent No. 5,008,752 discloses an interpolator for applying two different types of interpolation fiinctions depending on the type of image data present. This technique switches between different inte ⁇ olation functions when a difference between successive pixel values exceeds a predetermined threshold, indicating a transition from pictorial to text information.
  • United States Patent No. 4,468,693 discloses an inte ⁇ olation apparatus that allows a user to manually switch between cubic spline inte ⁇ olation, bilinear inte ⁇ olation, and pixel replication inte ⁇ olation according to the user's preference.
  • 327,257 discloses an apparatus that applies more than one inte ⁇ olation function to a set of image data to achieve a desired appearance.
  • This apparatus creates both a sha ⁇ inte ⁇ olated image and a smooth inte ⁇ olated image by applying a cubic spline function to a set of image data, and then creating a third image by combining the sha ⁇ and smooth inte ⁇ olated images with selected weighting coefficients to produce a sha ⁇ or smooth resultant image.
  • the above patent disclosures generally address the effectiveness of different types of inte ⁇ olation functions for different types of image data.
  • the effect of other image processing variables has not been adequately addressed.
  • the effectiveness of a particular inte ⁇ olation function may vary according to the degree of expansion required by the format of the image data on the imaging media. If a format requires a large image expansion factor for an image, it is, of course, necessary to create a greater number of intermediate pixels between adjacent original unexpanded pixels. As a result, there will be a higher chance of introducing imaging artifacts such as blurring. If the format selection calls for an smaller expansion rate, however, the blurring will be less perceptible. Hence, an inte ⁇ olation function considered inappropriate for use with larger expansion rates, due to high incidence of blurring, may nevertheless be appropriate for use with smaller expansion rates.
  • the response characteristics of different imaging media can greatly affect the visible results of the inte ⁇ olation. It has been observed, for example, that dry silver film has a development speed that is significantly slower than that of conventional silver halide film, in both toe and shoulder speeds. This slower development speed can render inte ⁇ olation functions previously considered effective for silver halide films less effective for dry silver film.
  • the slower development speed requires that the scanning laser transmit a higher exposing energy to the imaging media to achieve a desired maximum optical density. The higher exposing energy translates to a larger dynamic range for the scanning laser in order to expose every gradation on the imaging media.
  • the available dynamic range of existing, commercially available scanning lasers may be limited for required exposing wavelengths in the infra-red regions.
  • the scanning laser may be incapable of producing the desired number of gradations.
  • the slower development speed of the dry silver film given the limitations of the scanning laser, can result in loss of high spatial frequencies.
  • the visible manifestation of the loss of high spatial frequencies may be a loss of apparent edge sha ⁇ ness producing an undesirable blurring in high contrast areas.
  • the blurring is particularly apparent in those areas involving a transition from image data representing alphanumeric text at a minimum density to image data representing pictorial information at much higher densities approaching maximum density. Therefore, an inte ⁇ olation function considered appropriate for silver halide film may produce inadequate results with a dry silver film.
  • CRT monitor medium has several problems when compared to a print film for image presentation.
  • a CRT monitor suffers from severe dynamic range limitations (maximum intensity to minimum intensity range). relative to film.
  • the dynamic range of a CRT monitor limits perceptible image gradation presentation to about 100 - 150 levels as compared to print films, which may provide 1500-2000 levels of gradation.
  • CRT monitors may suffer blurring in high contrast areas.
  • inte ⁇ olation techniques generally fail to address the effect of format and/or media variations on inte ⁇ olation results. Consequently, image quality continues to be a concern for inte ⁇ olated images formed with different formats on different imaging media. Accordingly, there is a need for an inte ⁇ olation system capable of selecting inte ⁇ olation functions appropriate for particular formats and/or different imaging media.
  • the present invention is directed to a system and method for adaptive inte ⁇ olation of image data.
  • the system and method of the present invention are capable of selecting one of a plurality of different inte ⁇ olation kernel lookup tables based on a format selected for the image data.
  • the system and method of the present invention also are capable of selecting one of a plurality of different inte ⁇ olation kernel lookup tables based on an imaging media selected for presentation of the image data.
  • the system and method of the present invention enable application of an appropriate inte ⁇ olation kernel to the image data to improve visible results.
  • the system and method of the present invention are capable of selecting one of the inte ⁇ olation kernel lookup tables to achieve a particular appearance characteristic based on a degree of sha ⁇ ness or smoothness desired by a user.
  • the present invention provides, in a first embodiment, a system for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data representing inte ⁇ olated pixels, the system comprising media selection means for selecting one of a plurality of imaging media for presentation of the image, format selection means for selecting one of a plurality of formats for presentation of the image on the imaging media, appearance characteristic selection means for selecting one of a plurality of appearance characteristics of the presentation of the image on the imaging media, a memory storing a plurality of different inte ⁇ olation kernel lookup tables, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, inte ⁇ olation kernel selection means for selecting, based on a combination of the format selected by the format selection means, the media selected by the media selection means, and the appearance characteristic selected by the appearance characteristic selection means, one of the inte ⁇ olation kernel lookup tables stored in
  • the present invention provides a system for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data, the system comprising format selection means for selecting one of a plurality of formats for presentation of the image on imaging media, a memory storing a plurality of different inte ⁇ olation kernel lookup tables, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, inte ⁇ olation kernel selection means for selecting, based on the format selected by the format selection means, one of the inte ⁇ olation kernel lookup tables stored in the memory, and inte ⁇ olation means for applying the selected inte ⁇ olation kernel lookup table to the image data to produce the inte ⁇ olated image data.
  • the present invention provides a system for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data, the system comprising media selection means for selecting one of a plurality of imaging media for presentation of the image, a memory storing a plurality of different inte ⁇ olation kernel lookup tables, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, inte ⁇ olation kernel selection means for selecting, based on the media selected by the media selection means, one of the inte ⁇ olation kernel lookup tables stored in the memory, and inte ⁇ olation means for applying the selected inte ⁇ olation kernel lookup table to the image data to produce the inte ⁇ olated image data.
  • the present invention provides a method for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data representing inte ⁇ olated pixels, the method comprising the steps of selecting one of a plurality of imaging media for presentation of the image, selecting one of a plurality of formats for presentation of the image on the imaging media, selecting one of a plurality of appearance characteristics of the presentation of the image on the imaging media, selecting one of a plurality of different inte ⁇ olation kernel lookup tables based on a combination of the selected media, the selected format, and the selected appearance characteristic, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, and applying the inte ⁇ olation coefficients contained in the selected inte ⁇ olation kernel lookup table to the image data to produce the inte ⁇ olated image data.
  • the present invention provides a method for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data, the method comprising the steps of selecting one of a plurality of formats for presentation of the image on imaging media, selecting one of a plurality of different inte ⁇ olation kernel lookup tables based on the selected format, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, and inte ⁇ olation means for applying the selected inte ⁇ olation kernel lookup table to the image data to produce the inte ⁇ olated image data.
  • the present invention provides a method for performing inte ⁇ olation on image data representative of a plurality of pixels within an image to produce inte ⁇ olated image data, the method comprising the steps of selecting one of a plurality of imaging media for presentation of the image, selecting one of a plurality of different inte ⁇ olation kernel lookup tables based on the selected media, each of the inte ⁇ olation kernel lookup tables containing a plurality of inte ⁇ olation coefficients, and inte ⁇ olation means for applying the selected inte ⁇ olation kernel lookup table to the image data to produce the inte ⁇ olated image data.
  • FIG. 1 is a functional block diagram of a system for adaptive inte ⁇ olation of image data, in accordance with the present invention
  • Fig. 2 is a functional block diagram of an inte ⁇ olation processor for inco ⁇ oration in the system of Fig. 1, in accordance with the present invention
  • Fig. 3 is a flow diagram illustrating a method for adaptive inte ⁇ olation of image data, in accordance with the present invention.
  • Fig. 1 is a functional block diagram of a system 10 for adaptive inte ⁇ olation of digital image data, in accordance with the present invention.
  • system 10 includes a format selector 12, an appearance characteristic selector 14, an output media selector 16, an inte ⁇ olation kernel lookup table memory 18, an inte ⁇ olation processor 20, and a formatting processor 22.
  • the system 10 further includes one or more output media provided by a cathode ray tube (CRT) display monitor 24 and a digital laser imager 26.
  • the inte ⁇ olation processor 20 receives image data, as indicated by block 28, from an input imaging device.
  • the input imaging device may comprise, for example, a diagnostic device, such as a magnetic resonance (MR), computed tomography (CT), conventional radiography (X-ray), or ultrasound device.
  • MR magnetic resonance
  • CT computed tomography
  • X-ray conventional radiography
  • image data 28 may represent an image of a physiological object obtained for diagnostic pu ⁇ oses.
  • the image data 28 received by inte ⁇ olation processor 20 is representative of a plurality of pixels within the image.
  • the number of pixels represented by image data 28 can be significantly less than the output resolution of output imaging media used for visible presentation of the image.
  • the system 10 is configured to perform inte ⁇ olation on image data 28 to produce inte ⁇ olated image data representing the original pixels and inte ⁇ olated pixels.
  • the system 10 selects one of a plurality of inte ⁇ olation kernel lookup tables based on a combination of a selected format, a selected output media, and a selected appearance characteristic of the image.
  • the inte ⁇ olation kernel lookup tables contain inte ⁇ olation coefficients that are applied to the original pixels in image data 28 to produce inte ⁇ olated pixels, thereby transforming the spatial resolution of image data 28 to match the resolution of the output media.
  • the format selector 12 selects one of a plurality of formats for presentation of an image on imaging media.
  • the format selector 12 enables a system user to select a format representing a number of images to be formed on a single page of the output imaging media.
  • system 10 may provide formats for as many images as can be accommodated by a single page of media or, as a practical matter, as many images as can be viewed for useful diagnostic pu ⁇ oses.
  • page may refer to a sheet of photosensitive media such as film printed by laser imager 26, or a screen displayed by CRT display monitor 24.
  • the format selector 12 can be realized, for example, by a user interface associated with a display panel on laser imager 26 or by a manual control panel on laser imager 26.
  • the number of images specified by the selected format determines the size of each individual image to be formed on the imaging media.
  • the size of each image determines the amount of expansion necessary to match the resolution of image data 28 to that of the output media.
  • the appearance characteristic selector 14 selects one of a plurality of appearance characteristics of the presentation of the image on the imaging media.
  • the appearance characteristics may include a plurality of degrees of apparent sha ⁇ ness of the image relative to a degree of apparent smoothness, when viewed by a human observer.
  • the appearance characteristic selector 14 enables a user to select a degree of apparent edge contrast in the visible representation of the image when formed on the imaging media to provide a more pleasing or diagnostically useful appearance.
  • the format selector 12 can be realized, for example, by a user interface associated with a display panel on laser imager 26 or by a manual control panel on laser imager 26.
  • the media selector 16 selects one of a plurality of imaging media for visible presentation of the image represented by image data 28.
  • media selector 16 enables the user to select either CRT monitor 24 or laser imager 26, and also the type of imaging media printed by the laser imager.
  • laser imager 26 may be configured to print the image on either wet chemically processed silver halide film or thermographic dry silver film.
  • the laser imager 26 conceivably could be configured to print images on other types of thermographic film.
  • the media selector 16 can be realized, for example, by a user interface associated with a display panel on laser imager 26 or by a manual control panel on laser imager 26.
  • media selector 16 may be configured to respond automatically to media selection information sent with image data 28 from an input imaging device.
  • the inte ⁇ olation kernel lookup table memory 18 stores a plurality of different inte ⁇ olation kernel lookup tables.
  • Each of the inte ⁇ olation kernel lookup tables contains a plurality of inte ⁇ olation coefficients calculated according to particular inte ⁇ olation kernel functions.
  • the inte ⁇ olation kernel lookup tables may include, for example, at least one inte ⁇ olation kernel lookup table containing a plurality of two-point inte ⁇ olation coefficients and at least one inte ⁇ olation kernel lookup table containing a plurality of four-point inte ⁇ olation coefficients.
  • memory 18 may store a variety of inte ⁇ olation kernel lookup tables including an inte ⁇ olation kernel lookup table containing a plurality of pixel replication inte ⁇ olation coefficients, an inte ⁇ olation kernel lookup table containing a plurality of bilinear inte ⁇ olation coefficients, an inte ⁇ olation kernel lookup table containing a plurality of third order polynomial inte ⁇ olation coefficients, an inte ⁇ olation kernel lookup table containing a plurality of fifth order polynomial inte ⁇ olation coefficients, and one or more inte ⁇ olation kernel lookup tables containing a plurality of cubic spline inte ⁇ olation coefficients.
  • the inte ⁇ olation processor 20 selects one of the inte ⁇ olation kernel lookup tables stored in memory 18 based on a combination of the format selected by format selector 12, the media selected by media selector 16, and the appearance characteristic selected by appearance characteristic selector 14.
  • the particular inte ⁇ olation kernel lookup table selected by inte ⁇ olation processor 20 produces inte ⁇ olated image data having an optimum visible appearance, given the format, media, and appearance characteristic specified by the user.
  • the inte ⁇ olation processor 20 determines a sampling increment based on the amount of pixels required by the selected format and the amount of original pixels provided by image data 28.
  • the inte ⁇ olation processor 20 then applies inte ⁇ olation coefficients contained in the selected inte ⁇ olation kernel lookup table that correspond to the sampling increment to image data 28 to produce the inte ⁇ olated image data.
  • the format processor 22 receives the inte ⁇ olated image data produced by inte ⁇ olation processor 20, and prepares the inte ⁇ olated image data for presentation on a single page of output imaging media. Specifically, format processor 22 assembles the inte ⁇ olated image data produced for each of a number of images to be printed or displayed on a single page according to a format selected by the user. The format processor 22 assigns the inte ⁇ olated image data for each image to a particular area of the page, and then rasterizes the assembled image data for application to drive either display monitor 24 or laser imager 26.
  • Fig. 2 is a functional block diagram of the inte ⁇ olation processor 20 shown in Fig. 1, in accordance with the present invention.
  • processor 20 includes an input image data buffer 30, a row inte ⁇ olator 32, a column inte ⁇ olator 34, and an output image data buffer 36.
  • the input image data buffer 30 receives image data 28 from an input imaging device.
  • the row inte ⁇ olator 32 processes original pixel values along each row of image data 28 and inte ⁇ olates them based on inte ⁇ olation coefficients retrieved from the selected inte ⁇ olation kernel table, as indicated by block 38.
  • the row inte ⁇ olator 32 multiplies the original pixel values by corresponding inte ⁇ olation coefficients 38, selected according to the sampling rate, and sums the product to produce inte ⁇ olated row pixel values.
  • the resulting inte ⁇ olated row pixel values each correspond to an inte ⁇ olated pixel in the image, and represent an intensity of that inte ⁇ olated pixel.
  • the inte ⁇ olated row pixel values create additional columns of inte ⁇ olated pixels between columns of original pixels in image data 28.
  • the intensity of each inte ⁇ olated pixel is a function of the intensities of the original pixel values to which the inte ⁇ olation coefficients were applied, and the values of the inte ⁇ olation coefficients.
  • column inte ⁇ olator 34 processes them to produce inte ⁇ olated column pixel values. Specifically, column inte ⁇ olator 34 inte ⁇ olates both the original row pixels and the inte ⁇ olated row pixels along a column based on inte ⁇ olation coefficients 38 retrieved from the selected inte ⁇ olation kernel table. The column inte ⁇ olator 34 multiplies the original row pixels and the inte ⁇ olated row pixels along a column by corresponding inte ⁇ olation coefficients, and sums the product to produce inte ⁇ olated column pixel values.
  • the inte ⁇ olated column pixel values create additional rows of inte ⁇ olated pixels between rows of original pixels and inte ⁇ olated pixels.
  • the intensity of each inte ⁇ olated column pixel is a function of the intensities of the original column pixel values to which the inte ⁇ olation coefficients were applied, and the values of the inte ⁇ olation coefficients.
  • the row inte ⁇ olator 32 and column inte ⁇ olater 34 perform inte ⁇ olation of image data 28 in a well known manner.
  • the following FORTRAN code is provided, however, for pu ⁇ oses of illustration.
  • the code illustrates the computation involved in a four-point inte ⁇ olation process using a cubic spline inte ⁇ olation kernel function, a two-point inte ⁇ olation process using a fifth order polynomial function, and a two-point inte ⁇ olation process using bilinear inte ⁇ olation.
  • c ul, u2, u3, u4 are the calculated coefficients from c Cubic Spline kernel function c In this example, a 200% inte ⁇ olation is shown.
  • c c Two point inte ⁇ olation using Bilinear c
  • u2(i) l-cl
  • mbb(i,k) are the expanded image data
  • mnb(i,k) are the unexpanded original image data.
  • memory 18 stores inte ⁇ olation lookup tables containing inte ⁇ olation coefficients that represent the following inte ⁇ olation kernel functions:
  • Pixel Replication Kernel Function This two-point inte ⁇ olation kernel lookup table contains inte ⁇ olation coefficients representative of a conventional pixel replication inte ⁇ olation kernel function K ⁇ (x), whereby inte ⁇ olated pixels between two adjacent original pixels are simply assigned the intensity values of the closest original pixel.
  • This inte ⁇ olation kernel function can be expressed as:
  • K,(x) 0
  • x is a variable representing a location of the inte ⁇ olated pixel relative to the original pixels.
  • the pixel replication inte ⁇ olation kernel produces a sha ⁇ apparent edge contrast between pixels in the inte ⁇ olated image data.
  • This two-point inte ⁇ olation kernel lookup table contains inte ⁇ olation coefficients representative of a conventional bilinear inte ⁇ olation kernel function K (x), whereby inte ⁇ olated pixels between two adjacent original pixels are assigned linearly weighted intensity values based on distance from the closest adjacent original pixel.
  • 0 ⁇
  • This bilinear inte ⁇ olation kernel function produces a smoother apparent edge contrast between pixels in the inte ⁇ olated image data, relative to pixel replication.
  • This two-point inte ⁇ olation kernel lookup table contains inte ⁇ olation coefficients representative of a conventional third order polynomial kernel function K 3 (x), whereby inte ⁇ olated pixels between two adjacent original pixels are assigned intensity values based on the following expression:
  • This two-point inte ⁇ olation kernel lookup table contains inte ⁇ olation coefficients representative of a conventional fifth order polynomial kernel function K*(x), whereby inte ⁇ olation pixels between two adjacent original pixels are assigned intensity values based on the following expression:
  • K,(x) -6
  • ⁇ 1 K,(x) 0
  • the fifth order polynomial kernel function produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by both the third order polynomial inte ⁇ olation kernel function K 3 (x) and the bilinear replication inte ⁇ olation kernel function K (x), but smoother than that produced by the pixel replication inte ⁇ olation kernel function K ⁇ (x).
  • This four-point inte ⁇ olation kernel lookup table contains inte ⁇ olation coefficients representative of a conventional cubic spline kernel function Ks(x), whereby inte ⁇ olation pixels between two adjacent original pixels are assigned intensity values based on the following expression:
  • K 5 (x) (a + 2)
  • 2 + 1 0 ⁇
  • 1
  • Kj(x)
  • Cubic Spline Function (a -1.0).
  • the inte ⁇ olation kernel function K l0 (x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function K ⁇ x).
  • Cubic Spline Function ( - 1.2).
  • This four-point inte ⁇ olation kernel lookup table corresponds to the above expression K 5 (x) with the coefficient a - -1.2 to realize an inte ⁇ olation kernel function K ⁇ (x).
  • the inte ⁇ olation kernel function Kn(x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function K ⁇ 0 (x).
  • Cubic Spline Function ( -1.4).
  • the inte ⁇ olation kernel function K !2 (x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function Ku(x).
  • Cubic Spline Function (a -1.6).
  • the inte ⁇ olation kernel function K_ (x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function K ⁇ 2 (x).
  • Cubic Spline Function (a -1.8).
  • the inte ⁇ olation kernel function Ku(x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function K ⁇ 3 (x).
  • Cubic Spline Function (a -2.0).
  • the inte ⁇ olation kernel function K ⁇ s(x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function Ku(x).
  • Cubic Spline Function (a -2.2).
  • the inte ⁇ olation kernel function K ⁇ 6 (x) produces an apparent edge contrast between pixels in the inte ⁇ olated image that is sha ⁇ er than that produced by the inte ⁇ olation kernel function K ⁇ s(x).
  • the inte ⁇ olation processor 20 selects one of the above inte ⁇ olation kernel lookup tables K ⁇ -K ]6 to improve the appearance of a visible representation of the inte ⁇ olated image data based on the selected format, the selected media, and the selected appearance characteristic.
  • inte ⁇ olation processor 20 selects from memory 18 the best available inte ⁇ olation kernel function K Ki 6 for the present conditions.
  • inte ⁇ olation processor 20 first narrows the field of inte ⁇ olation kernel lookup tables stored in memory 18 according to the format selected by format selector 12. For selected formats requiring smaller expansion rates, inte ⁇ olation processor 20 selects one of the two-point inte ⁇ olation kernel lookup tables K1-K , in view of the lower incidence of blurring with small-scale expansion. Because the smaller expansion rate produces less blurring, a two-point inte ⁇ olation kernel lookup table can provide acceptable visible results. In this manner, the selection of a two-point, rather than four-point, inte ⁇ olation kernel lookup table by inte ⁇ olation processor 20 reduces the amount of computational effort necessary to produce the inte ⁇ olated image data.
  • inte ⁇ olation processor 20 selects one of the four-point inte ⁇ olation kernel tables K-rK ⁇ 6 .
  • the inte ⁇ olation processor 20 thereby produces better spatial frequency when needed to avoid the higher incidence of blurring that can occur with larger expansion.
  • the four-point inte ⁇ olation kernel requires added computation, but is necessary to ensure a pleasing or diagnostically useful visible presentation of the inte ⁇ olated image data on the output media.
  • inte ⁇ olation processor 20 As an example of the criteria applied by inte ⁇ olation processor 20 to select a type of inte ⁇ olation kernel lookup table, satisfactory visible results have been observed for the use of two-point inte ⁇ olation kernel tables with selected formats requiring expansion rates that do not exceed a factor of three in both horizontal and vertical directions. For selected formats that require expansion rates of greater than a factor of three in both horizontal and vertical directions, inte ⁇ olation processor 20 should be configured to select one of the four-point inte ⁇ olation kernel tables. At expansion rates of less than a factor of three, the degree of blurring that may occur with a two-point inte ⁇ olation kernel has been found to be acceptable, particularly in view of the computational savings gained.
  • inte ⁇ olation processor 20 After inte ⁇ olation processor 20 has narrowed the selection process to either the set of two-point inte ⁇ olation kernel lookup tables or the set of four- point inte ⁇ olation kernel lookup tables, the process turns to selection of a subset of lookup tables appropriate for the type of output imaging media selected by output media selector 16.
  • Both the set of two-point inte ⁇ olation lookup tables K1-K4 and the set of four-point inte ⁇ olation kernel lookup tables Ks-K ]6 include inte ⁇ olation coefficients representing inte ⁇ olation kernel functions designed to produce different appearance characteristics, i.e., different degrees of sha ⁇ ness or smoothness.
  • the set of two-point inte ⁇ olation kernel tables includes, in descending order of apparent sha ⁇ ness, the pixel replication inte ⁇ olation kernel table Ki, the fifth order polynomial inte ⁇ olation kernel table Id, the third order polynomial inte ⁇ olation kernel table K 3 , and the bilinear inte ⁇ olation kernel table K 2 .
  • the set of four-point inte ⁇ olation kernel tables includes, in descending order of apparent sha ⁇ ness, cubic spline inte ⁇ olation kernel tables K 5 -K ⁇ 6 modified by the value of coefficient a to adjust sha ⁇ ness.
  • the inte ⁇ olation processor 20 selects a subset of appropriate inte ⁇ olation kernel lookup tables by considering the response characteristics of the output imaging media, which can greatly affect the visible results of the inte ⁇ olation.
  • Thermographic dry silver film has a development speed that is significantly slower than that of conventional silver halide film, in both toe and shoulder speeds.
  • the scanning laser in laser imager 26 may have an insufficient dynamic range to compensate for the slower development speed. Consequently, blurring can result in high contrast areas of the visible presentation of the inte ⁇ olated image data, such as those areas involving a transition from image data representing alphanumeric text at a minimum density to image data representing pictorial information at much higher densities approaching maximum density.
  • a CRT display monitor suffers from severe dynamic range limitations, and therefore can also exhibit blurring in high contrast areas.
  • inte ⁇ olation processor 20 selects one of the inte ⁇ olation kernel lookup tables Kj-Ki ⁇ producing a relatively smooth apparent edge contrast.
  • the inte ⁇ olation processor 20 selects an inte ⁇ olation kernel lookup table producing a relatively smooth apparent edge contrast in recognition of the faster development speed of the silver halide film, and hence the smaller dynamic range required to reproduce the desired optical densities.
  • inte ⁇ olation processor 20 selects one of two-point inte ⁇ olation kernel tables Ki-I providing a relatively smooth apparent edge contrast, such as the bilinear inte ⁇ olation kernel table K 2 .
  • inte ⁇ olation processor 20 selects one of the four-point inte ⁇ olation kernel tables K 3 -Ki6 providing a relatively smooth apparent edge contrast.
  • the sixth four-point inte ⁇ olation kernel table K ⁇ with coefficient a set to -0.2, provides a particularly smooth apparent edge contrast when applied to the inte ⁇ olated image data.
  • inte ⁇ olation processor 20 selects one of the inte ⁇ olation kernel lookup tables Ki-Ku producing a sha ⁇ er apparent edge contrast than that for silver halide film.
  • the inte ⁇ olation processor 20 selects an inte ⁇ olation kernel lookup table producing a relatively sha ⁇ appearance in recognition of the slower development speed of the dry silver film, and hence the larger dynamic range required to reproduce the desired optical densities.
  • the sha ⁇ er inte ⁇ olation kernel lookup table provides a mechanism for improving the visible appearance of the presentation of the inte ⁇ olated image data on the dry silver film.
  • the sha ⁇ ness produced by the selected inte ⁇ olation kernel lookup table produces an "overshoot undershoot" effect that compensates for the slower development speed of the dry silver film.
  • the overshoot and undershoot are clipped in high contrast areas by the minimum density Dmin and the maximum density Dmax.
  • the sha ⁇ ness of the selected inte ⁇ olating kernel lookup table increases the apparent spatial frequency of the inte ⁇ olated image data by increasing the apparent slope between adjacent inte ⁇ olated pixels.
  • the selected inte ⁇ olating kernel lookup table increases the apparent slope between minimum density Dmin and maximum density Dmax in high contrast areas, such as those involving a transition from image data representing alphanumeric text at a minimum density to image data representing pictorial information at much higher densities approaching maximum density.
  • the other areas of the visible presentation having varying densities also will be affected by the sha ⁇ er inte ⁇ olation kernel lookup table. The effect is not readily apparent, however, for inte ⁇ olation with formats requiring smaller expansion factors.
  • the four- point inte ⁇ olation kernel lookup tables selected for formats requiring larger expansion factors conceal the effect by avoiding the introduction of a large amount of high spatial frequency artifact and aliasing.
  • inte ⁇ olation processor 20 selects one of the two-point inte ⁇ olation kernel tables K1-K4 providing a relatively sha ⁇ apparent edge contrast, such as the pixel replication inte ⁇ olation kernel table K t , the third order polynomial kernel table K , or the fifth order polynomial kernel table t.
  • inte ⁇ olation processor 20 selects one of four-point inte ⁇ olation kernel tables K 5 -K ⁇ 6 providing relatively sha ⁇ er apparent edge contrasts.
  • the inte ⁇ olation kernel table K ⁇ provides a very high degree of overshoot and undershoot, producing a very sha ⁇ apparent edge contrast when applied to the inte ⁇ olated image data.
  • the sha ⁇ er inte ⁇ olation kernel lookup table provides the benefits of higher apparent slope between Dmin and Dmax, compensating for the slow development speed of the dry silver film, but does not suffer significantly from the overshoot and undershoot effect, which would ordinarily be undesirable.
  • inte ⁇ olation processor 20 selects one of the inte ⁇ olation kernel lookup tables Ki-Ki ⁇ producing an even sha ⁇ er apparent edge contrast than that required by the dry silver film.
  • the inte ⁇ olation processor 20 selects an inte ⁇ olation kernel lookup table producing a relatively sha ⁇ appearance in recognition of the severely limited dynamic range of the CRT monitor.
  • the sha ⁇ er inte ⁇ olation kernel lookup table provides a mechanism for improving the visible appearance of the presentation of the inte ⁇ olated image data on the CRT monitor.
  • the sha ⁇ ness produced by the selected inte ⁇ olation kernel lookup table produces the overshoot and undershoot effect to provide a higher apparent slope between maximum intensity and minimum intensity on the monitor.
  • the overshoot/undershoot effect compensates for the limited dynamic range of the CRT monitor, but is clipped in high contrast areas by the minimum and maximum intensities.
  • the sha ⁇ ness of the selected inte ⁇ olating kernel lookup table increases the apparent spatial frequency of the inte ⁇ olated image data by increasing the apparent slope between adjacent inte ⁇ olated pixels in high contrast areas, such as those involving a transition from image data representing alphanumeric text at a minimum intensity to image data representing pictorial information at much higher intensities approaching maximum intensity.
  • inte ⁇ olation processor 20 selects one of the two-point inte ⁇ olation kernel tables K K_ ⁇ providing a sha ⁇ apparent edge contrast.
  • the pixel replication inte ⁇ olation kernel table Ki and the fifth order polynomial kernel table K «, in particular, produce inte ⁇ olated image data having sha ⁇ er apparent edge contrasts.
  • inte ⁇ olation processor 20 selects one of the four-point inte ⁇ olation kernel tables K 5 -K ⁇ 6 providing relatively sha ⁇ er apparent edge contrasts.
  • the inte ⁇ olation kernel table Ki ⁇ provides a maximum degree of overshoot and undershoot, producing a very sha ⁇ apparent edge contrast when applied to the inte ⁇ olated image data.
  • the sha ⁇ er inte ⁇ olation kernel lookup table provides a higher apparent slope to compensate for the limited dynamic range of CRT monitor 24, thereby improving the visible presentation of the inte ⁇ olated image data to a human viewer.
  • inte ⁇ olation processor 20 After inte ⁇ olation processor 20 has narrowed the selection process to either the set of two-point inte ⁇ olation kernel lookup tables K 1 -K or the set of four-point inte ⁇ olation kernel lookup tables K5-K16 based on the selected format, and to a particular subset of the inte ⁇ olation kernel lookup tables Ki-Ki ⁇ based on the selected media the process turns to selection of a particular inte ⁇ olation kernel lookup table based on the appearance characteristic selected by appearance characteristic selector 14.
  • the selection by inte ⁇ olation processor 20 is largely driven by format and media considerations, it still may be desirable to allow a system user some latitude for sha ⁇ ness and smoothness.
  • inte ⁇ olation processor 20 selects one of the four-point inte ⁇ olation kernel tables K5-K16 providing a sha ⁇ apparent edge contrast.
  • the apparent sha ⁇ ness produced by each of inte ⁇ olation kernel tables K5-K16 increases in response to an increase in the coefficient a.
  • inte ⁇ olation processor 20 selects a higher-numbered inte ⁇ olation kernel lookup table having a coefficient a that corresponds to the desired sha ⁇ ness.
  • Fig. 3 is a flow diagram illustrating a method for adaptive inte ⁇ olation of image data, in accordance with the present invention. It will be recognized that the method of the present invention can be readily implemented by the system 10 described above. As shown in the flow diagram of Fig. 3, the method includes a step of selecting one of a plurality of formats for presentation of an image on an imaging media, as indicated by block 40. A particular imaging media is selected for presentation of an image, as indicated by block 42. Further, one of a plurality of appearance characteristics, representing sha ⁇ ness or smoothness, are selected for the presentation of the image on the imaging media, as indicated by block 44.
  • one of a plurality of different inte ⁇ olation kernel lookup tables then is selected based on a combination of the selected format, the selected media, and the selected appearance characteristic.
  • Each of inte ⁇ olation kernel lookup tables contains a plurality of inte ⁇ olation coefficients representing the inte ⁇ olation carried by a different inte ⁇ olation kernel functions, as described above with respect to system 10.
  • the inte ⁇ olation coefficients contained in the selected inte ⁇ olation kernel lookup table are applied to the image data to produce the inte ⁇ olated image data.
  • the following example is provided to illustrate a system and method for adaptive inte ⁇ olation of image data, in accordance with the present invention.
  • the following example illustrates one possible implementation of the selection process carried out by the system and method of the present invention.
  • one of inte ⁇ olation kernel lookup tables Ki-K* is selected if the selected format requires an expansion factor of less than or equal to three in horizontal and vertical directions
  • one of inte ⁇ olation kernel lookup tables K 5 -K ⁇ is selected if the selected format requires an expansion factor of greater than three in horizontal and vertical direction.
  • one of inte ⁇ olation kernel lookup tables K 1 -K4 is selected according to the selected appearance characteristic, with Ki representing the sha ⁇ est appearance and K 2 the smoothest. If the selected media is silver halide and the selected format requires an expansion factor of greater than three, one of inte ⁇ olation kernel lookup tables in the subset K 5 -K 12 is selected according to the selected appearance characteristic, with K5 representing the smoothest appearance in the subset and K ⁇ 2 the sha ⁇ est.
  • one of inte ⁇ olation kernel lookup tables K 1 -K- 1 is selected according to the selected appearance characteristic, with K] representing the sha ⁇ est appearance and K 2 the smoothest. If the selected media is dry silver and the selected format requires an expansion factor of greater than three, one of inte ⁇ olation kernel lookup tables in the subset Kg-K ⁇ is selected according to the selected appearance characteristic, with Kg representing the least sha ⁇ appearance in the subset and K ⁇ 4 the sha ⁇ est.
  • one of inte ⁇ olation kernel lookup tables K 1 -K4 is selected according to the selected appearance characteristic, with Ki representing the sha ⁇ est appearance and K the smoothest. If the selected media is CRT monitor 24 and the selected format requires an expansion factor of greater than three, one of inte ⁇ olation kernel lookup tables in the subset Kio-Ki ⁇ is selected according to the selected appearance characteristic, with K10 representing the least sha ⁇ appearance in the subset and K J6 the sha ⁇ est.

Abstract

La présente invention concerne un système et un procédé d'interpolation adaptative de données vidéo, qui choisissent une fonction d'interpolation du système d'exploitation parmi plusieurs différentes, fondées sur une ou plusieurs variables, comme un format sélectionné pour la présentation visible des données vidéo interpolées sur un média d'imagerie de sortie, un type sélectionné de média d'imagerie de sortie pour la présentation visible des données vidéo interpolées et une caractéristique d'aspect, telle que la netteté ou la douceur, de la présentation visible des données vidéo interpolées. Le système et le procédé choisissent une fonction d'interpolation appropriée du système d'exploitation pour améliorer l'aspect de la présentation visible des données vidéo interpolées sur le média d'imagerie de sortie.
PCT/US1995/013560 1994-11-23 1995-10-10 Systeme et procede d'interpolation adaptative de donnees video WO1996016380A1 (fr)

Priority Applications (3)

Application Number Priority Date Filing Date Title
JP8516854A JPH10509824A (ja) 1994-11-23 1995-10-10 画像データの適応的補間のためのシステム及び方法
EP95938826A EP0793836A1 (fr) 1994-11-23 1995-10-10 Systeme et procede d'interpolation adaptative de donnees video
AU40062/95A AU4006295A (en) 1994-11-23 1995-10-10 System and method for adaptive interpolation of image data

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
US34456194A 1994-11-23 1994-11-23
US08/344,561 1994-11-23

Publications (1)

Publication Number Publication Date
WO1996016380A1 true WO1996016380A1 (fr) 1996-05-30

Family

ID=23351042

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/US1995/013560 WO1996016380A1 (fr) 1994-11-23 1995-10-10 Systeme et procede d'interpolation adaptative de donnees video

Country Status (5)

Country Link
US (1) US5774601A (fr)
EP (1) EP0793836A1 (fr)
JP (1) JPH10509824A (fr)
AU (1) AU4006295A (fr)
WO (1) WO1996016380A1 (fr)

Cited By (21)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2314477A (en) * 1996-06-19 1997-12-24 Quantel Ltd Image magnification processing system employing non-linear interpolation
EP0874330A2 (fr) * 1997-04-21 1998-10-28 Hewlett-Packard Company Amélioration d'images utilisant une interpolation de surface
EP0881595A1 (fr) * 1997-05-30 1998-12-02 Fujifilm Electronic Imaging Limited Appareil et procédé de génération d'images en mode point
EP0895191A2 (fr) * 1997-07-31 1999-02-03 Sony Corporation Appareil et procédé de traitement hiérarchique d'image
WO1999028864A1 (fr) * 1997-11-28 1999-06-10 Oce Printing Systems Gmbh Procede de conversion de donnees numeriques comprises dans une trame d'une premiere definition en donnees cibles numeriques d'une seconde definition
WO1999041691A1 (fr) * 1998-02-13 1999-08-19 Eastman Kodak Company Procede et systeme pour visualiser et configurer a distance les donnees de sortie d'un dispositif d'imagerie medical
GB2334845A (en) * 1998-02-28 1999-09-01 Samsung Electronics Co Ltd A scan format converter using look-up tables for converting a video format by bi-sigmoidal interpolation
EP0949585A2 (fr) * 1998-04-06 1999-10-13 Seiko Epson Corporation Dispositif et méthode pour interpoler des données d'image
EP0949586A2 (fr) * 1998-04-06 1999-10-13 Seiko Epson Corporation Dispositif et méthode d'interpolation de données d'image
EP0999696A1 (fr) * 1998-04-20 2000-05-10 Seiko Epson Corporation Support d'enregistrement de programmes de commande d'impression, controleur d'impression, et procede de commande de l'impression
AU729482B2 (en) * 1998-12-18 2001-02-01 Canon Kabushiki Kaisha Image interpolation with a continuous two-dimensional kernel
AU730522B1 (en) * 1998-12-18 2001-03-08 Canon Kabushiki Kaisha Continuous kernel image interpolation
GB2355377A (en) * 1999-10-11 2001-04-18 Ctx Opto Electronics Corp Digital image scaling with automatic image quality adjustment
US6272469B1 (en) * 1998-11-25 2001-08-07 Ge Medical Systems Global Technology Company, Llc Imaging system protocol handling method and apparatus
AU745082B2 (en) * 1998-12-18 2002-03-14 Canon Kabushiki Kaisha A steerable kernel for image interpolation
AU745562B2 (en) * 1998-12-18 2002-03-21 Canon Kabushiki Kaisha A method of kernel selection for image interpolation
AU748831B2 (en) * 1998-12-18 2002-06-13 Canon Kabushiki Kaisha A modified kernel for image interpolation
EP1321896A2 (fr) * 2001-12-21 2003-06-25 International Business Machines Corporation Procédé et circuits de mise à l'échelle d'images utilisant des réseaux neuronaux
US6714210B1 (en) 1998-12-18 2004-03-30 Canon Kabushiki Kaisha Continuous kernel image interpolation
US7054507B1 (en) 1998-12-18 2006-05-30 Canon Kabushiki Kaisha Method of kernel selection for image interpolation
US7352918B2 (en) 2001-12-21 2008-04-01 International Business Machines Corporation Method and circuits for scaling images using neural networks

Families Citing this family (38)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0753828B1 (fr) * 1995-07-13 2002-10-02 Fuji Photo Film Co., Ltd. Procédé et appareil pour l'interpolation des signaux d'images
JP3753197B2 (ja) 1996-03-28 2006-03-08 富士写真フイルム株式会社 画像データの補間演算方法およびその方法を実施する装置
US7057653B1 (en) * 1997-06-19 2006-06-06 Minolta Co., Ltd. Apparatus capable of image capturing
EP0921494B1 (fr) * 1997-12-03 2003-10-15 Fuji Photo Film Co., Ltd. Procédé et appareil pour l'interpolation des signaux d'image
US6801339B1 (en) * 1998-03-26 2004-10-05 Fuji Photo Film Co., Ltd. Image processing method and apparatus
US5991464A (en) * 1998-04-03 1999-11-23 Odyssey Technologies Method and system for adaptive video image resolution enhancement
US6721694B1 (en) * 1998-10-13 2004-04-13 Raytheon Company Method and system for representing the depths of the floors of the oceans
AUPP779798A0 (en) * 1998-12-18 1999-01-21 Canon Kabushiki Kaisha A modified kernel for image interpolation
AUPQ377899A0 (en) * 1999-10-29 1999-11-25 Canon Kabushiki Kaisha Phase three kernel selection
US7405740B1 (en) * 2000-03-27 2008-07-29 Stmicroelectronics, Inc. Context sensitive scaling device and method
ATE377812T1 (de) * 2000-04-24 2007-11-15 Seiko Epson Corp Vorrichtung und verfahren zur interpolation von bilddaten
FR2810131B1 (fr) * 2000-06-08 2005-04-08 Stg Interactive Systeme de publication de donnees multimedias
US6647137B1 (en) 2000-07-10 2003-11-11 International Business Machines Corporation Characterizing kernel function in photolithography based on photoresist pattern
JP4465570B2 (ja) * 2000-09-26 2010-05-19 ソニー株式会社 画像処理装置および方法、並びに記録媒体
US6915024B1 (en) * 2000-09-29 2005-07-05 Hewlett-Packard Development Company, L.P. Image sharpening by variable contrast mapping
US7453066B2 (en) * 2001-01-06 2008-11-18 Koninklijke Philips Electronics N.V. Method and apparatus to recover a dead pixel in digital imaging systems
JP2002262094A (ja) * 2001-02-27 2002-09-13 Konica Corp 画像処理方法及び画像処理装置
US6529574B1 (en) 2001-07-18 2003-03-04 Ge Medical Systems Global Technology Company, Llc Methods and apparatus for FOV-dependent aliasing artifact reduction
US6587537B1 (en) * 2002-04-01 2003-07-01 Ge Medical Systems Global Technology Company, Llc Methods and apparatus for multi-slice image reconstruction
US7355757B2 (en) * 2002-10-01 2008-04-08 Seiko Epson Corporation Fast edge reconstruction with upscaling for pulse width modulation rendering
US7096132B2 (en) * 2002-10-17 2006-08-22 Qualcomm Incorporated Procedure for estimating a parameter of a local maxima or minima of a function
US6873910B2 (en) 2002-10-22 2005-03-29 Qualcomm Incorporated Procedure for searching for position determination signals using a plurality of search modes
US7254282B2 (en) * 2003-01-27 2007-08-07 Scimed Life Systems, Inc. Systems and methods for transforming video images using fast visual interpolation
US7391933B2 (en) * 2003-10-30 2008-06-24 Samsung Electronics Co., Ltd. Method and apparatus for image interpolation based on adaptive polyphase filters
JP3833212B2 (ja) * 2003-11-19 2006-10-11 シャープ株式会社 画像処理装置、画像処理プログラムおよび可読記録媒体
US7315661B2 (en) * 2004-04-01 2008-01-01 Mediatek Inc. Directional interpolation method using DCT information and related device
US7720303B2 (en) * 2004-04-28 2010-05-18 Hewlett-Packard Development Company, L.P. Polynomial approximation based image filter methods, systems, and machine-readable media
US7496242B2 (en) * 2004-12-16 2009-02-24 Agfa Inc. System and method for image transformation
GB0500805D0 (en) * 2005-01-14 2005-02-23 Snell & Wilcox Ltd Image processing
JP4375269B2 (ja) * 2005-03-29 2009-12-02 セイコーエプソン株式会社 印刷制御方法、印刷制御装置、及び印刷制御プログラム
US7773831B2 (en) * 2005-05-10 2010-08-10 Agfa-Gevaert N.V. Method for visual inspection of size-reduced digital image
KR100754735B1 (ko) * 2006-02-08 2007-09-03 삼성전자주식회사 에지 신호 성분을 이용한 효율적인 영상 확대 방법 및 이를위한 장치
US7469259B2 (en) * 2006-08-10 2008-12-23 Kabushiki Kaisha Toshiba System and method for employing an extended boundary lookup table
US20080111823A1 (en) * 2006-11-13 2008-05-15 Faraday Technology Corp. Graphics processing system
JP4905187B2 (ja) * 2007-03-09 2012-03-28 ソニー株式会社 画像処理装置、撮像装置、および画像処理方法、並びにコンピュータ・プログラム
JP4444354B2 (ja) 2008-08-04 2010-03-31 株式会社東芝 画像処理装置、および画像処理方法
KR101544914B1 (ko) * 2009-02-12 2015-08-17 삼성전자주식회사 보간 픽셀을 생성하는 방법
US20110298972A1 (en) * 2010-06-04 2011-12-08 Stmicroelectronics Asia Pacific Pte. Ltd. System and process for image rescaling using adaptive interpolation kernel with sharpness and de-ringing control

Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4468693A (en) * 1981-07-14 1984-08-28 Dai Nippon Printing Co., Ltd. Video printing apparatus
EP0244141A2 (fr) * 1986-04-21 1987-11-04 Konica Corporation Système de traitement d'image
WO1990016035A2 (fr) * 1989-06-16 1990-12-27 Eastman Kodak Company Interpolateur a image numerique
US5202670A (en) * 1986-08-20 1993-04-13 Canon Kabushiki Kaisha Image processing apparatus
US5327257A (en) * 1992-02-26 1994-07-05 Cymbolic Sciences International Ltd. Method and apparatus for adaptively interpolating a digital image

Family Cites Families (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4595958A (en) * 1984-08-24 1986-06-17 Minnesota Mining And Manufacturing Company Multiformat image recordation
US4583128A (en) * 1984-08-27 1986-04-15 Minnesota Mining And Manufacturing Company Continuous tone recording system incorporating feedback control circuit
JPH0786924B2 (ja) * 1989-04-20 1995-09-20 富士写真フイルム株式会社 画像拡大縮小方法および装置
US5008752A (en) * 1989-06-16 1991-04-16 Eastman Kodak Company Digital image interpolator with multiple interpolation algorithms
JPH05504433A (ja) * 1990-12-31 1993-07-08 イー・アイ・デュポン・ドゥ・ヌムール・アンド・カンパニー 画像のサイズを変換する方法、フィルタを設計する方法および出力画像空間のピクセルをサイズ変換された入力画像空間に写像する方法
US5319719A (en) * 1991-05-15 1994-06-07 Konica Corporation Processing apparatus for radiographic image signals
JPH0514654A (ja) * 1991-07-04 1993-01-22 Kowa Co 画像信号処理装置
US5369499A (en) * 1991-11-13 1994-11-29 Eastman Kodak Company Calibration method for video image reproduction with electronic printer and video monitor

Patent Citations (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4468693A (en) * 1981-07-14 1984-08-28 Dai Nippon Printing Co., Ltd. Video printing apparatus
EP0244141A2 (fr) * 1986-04-21 1987-11-04 Konica Corporation Système de traitement d'image
US5202670A (en) * 1986-08-20 1993-04-13 Canon Kabushiki Kaisha Image processing apparatus
WO1990016035A2 (fr) * 1989-06-16 1990-12-27 Eastman Kodak Company Interpolateur a image numerique
US5327257A (en) * 1992-02-26 1994-07-05 Cymbolic Sciences International Ltd. Method and apparatus for adaptively interpolating a digital image

Cited By (38)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
GB2314477A (en) * 1996-06-19 1997-12-24 Quantel Ltd Image magnification processing system employing non-linear interpolation
EP0874330A2 (fr) * 1997-04-21 1998-10-28 Hewlett-Packard Company Amélioration d'images utilisant une interpolation de surface
EP0874330A3 (fr) * 1997-04-21 2000-05-03 Hewlett-Packard Company Amélioration d'images utilisant une interpolation de surface
EP0881595A1 (fr) * 1997-05-30 1998-12-02 Fujifilm Electronic Imaging Limited Appareil et procédé de génération d'images en mode point
US6307972B1 (en) 1997-05-30 2001-10-23 Fujifilm Electronic Imaging Ltd Method and apparatus for generating a bitmap
EP0895191A3 (fr) * 1997-07-31 2000-06-21 Sony Corporation Appareil et procédé de traitement hiérarchique d'image
EP0895191A2 (fr) * 1997-07-31 1999-02-03 Sony Corporation Appareil et procédé de traitement hiérarchique d'image
US6404924B1 (en) 1997-07-31 2002-06-11 Sony Corporation Image processing apparatus and method
WO1999028864A1 (fr) * 1997-11-28 1999-06-10 Oce Printing Systems Gmbh Procede de conversion de donnees numeriques comprises dans une trame d'une premiere definition en donnees cibles numeriques d'une seconde definition
US6836571B1 (en) 1997-11-28 2004-12-28 OCé PRINTING SYSTEMS GMBH Method for converting digital raster data of a first resolution into digital target data of a second resolution
WO1999041691A1 (fr) * 1998-02-13 1999-08-19 Eastman Kodak Company Procede et systeme pour visualiser et configurer a distance les donnees de sortie d'un dispositif d'imagerie medical
US6101407A (en) * 1998-02-13 2000-08-08 Eastman Kodak Company Method and system for remotely viewing and configuring output from a medical imaging device
GB2334845A (en) * 1998-02-28 1999-09-01 Samsung Electronics Co Ltd A scan format converter using look-up tables for converting a video format by bi-sigmoidal interpolation
GB2334845B (en) * 1998-02-28 2000-04-26 Samsung Electronics Co Ltd Method for making look-up tables for video format conversion and scan format converter using the look-up tables
EP0949586A3 (fr) * 1998-04-06 2001-10-10 Seiko Epson Corporation Dispositif et méthode d'interpolation de données d'image
EP0949586A2 (fr) * 1998-04-06 1999-10-13 Seiko Epson Corporation Dispositif et méthode d'interpolation de données d'image
EP0949585A2 (fr) * 1998-04-06 1999-10-13 Seiko Epson Corporation Dispositif et méthode pour interpoler des données d'image
US6760489B1 (en) 1998-04-06 2004-07-06 Seiko Epson Corporation Apparatus and method for image data interpolation and medium on which image data interpolation program is recorded
EP1369818A1 (fr) * 1998-04-06 2003-12-10 Seiko Epson Corporation Dispositif et méthode pour interpoler des données d'image
EP0949585A3 (fr) * 1998-04-06 2001-10-17 Seiko Epson Corporation Dispositif et méthode pour interpoler des données d'image
US6614553B2 (en) 1998-04-06 2003-09-02 Seiko Epson Corporation Apparatus and method for image data interpolation and medium on which image data interpolation program is recorded
US6510254B1 (en) 1998-04-06 2003-01-21 Seiko Epson Corporation Apparatus and method for image data interpolation and medium on which image data interpolation program is recorded
EP0999696A1 (fr) * 1998-04-20 2000-05-10 Seiko Epson Corporation Support d'enregistrement de programmes de commande d'impression, controleur d'impression, et procede de commande de l'impression
US6768559B1 (en) 1998-04-20 2004-07-27 Seiko Epson Corporation Medium on which printing control program is recorded, printing controller, and printing controlling method
EP0999696A4 (fr) * 1998-04-20 2003-06-04 Seiko Epson Corp Support d'enregistrement de programmes de commande d'impression, controleur d'impression, et procede de commande de l'impression
US6272469B1 (en) * 1998-11-25 2001-08-07 Ge Medical Systems Global Technology Company, Llc Imaging system protocol handling method and apparatus
AU730522B1 (en) * 1998-12-18 2001-03-08 Canon Kabushiki Kaisha Continuous kernel image interpolation
AU745082B2 (en) * 1998-12-18 2002-03-14 Canon Kabushiki Kaisha A steerable kernel for image interpolation
AU729482B2 (en) * 1998-12-18 2001-02-01 Canon Kabushiki Kaisha Image interpolation with a continuous two-dimensional kernel
US6714210B1 (en) 1998-12-18 2004-03-30 Canon Kabushiki Kaisha Continuous kernel image interpolation
AU748831B2 (en) * 1998-12-18 2002-06-13 Canon Kabushiki Kaisha A modified kernel for image interpolation
AU745562B2 (en) * 1998-12-18 2002-03-21 Canon Kabushiki Kaisha A method of kernel selection for image interpolation
US7054507B1 (en) 1998-12-18 2006-05-30 Canon Kabushiki Kaisha Method of kernel selection for image interpolation
GB2355377A (en) * 1999-10-11 2001-04-18 Ctx Opto Electronics Corp Digital image scaling with automatic image quality adjustment
EP1321896A2 (fr) * 2001-12-21 2003-06-25 International Business Machines Corporation Procédé et circuits de mise à l'échelle d'images utilisant des réseaux neuronaux
EP1321896A3 (fr) * 2001-12-21 2005-11-16 International Business Machines Corporation Procédé et circuits de mise à l'échelle d'images utilisant des réseaux neuronaux
US7352918B2 (en) 2001-12-21 2008-04-01 International Business Machines Corporation Method and circuits for scaling images using neural networks
US7734117B2 (en) 2001-12-21 2010-06-08 International Business Machines Corporation System for scaling images using neural networks

Also Published As

Publication number Publication date
AU4006295A (en) 1996-06-17
EP0793836A1 (fr) 1997-09-10
US5774601A (en) 1998-06-30
JPH10509824A (ja) 1998-09-22

Similar Documents

Publication Publication Date Title
US5774601A (en) System and method for adaptive interpolation of image data
US5131057A (en) Method for video-to-printing image resolution conversion
US6782137B1 (en) Digital image display improvement system and method
US5351314A (en) Method and apparatus for image enhancement using intensity dependent spread filtering
US20070086671A1 (en) Image processing apparatus
EP0632914B1 (fr) Procede et appareil pour realiser d'une maniere adaptative l'interpolation d'une image numerique
EP1032196B1 (fr) Méthode et appareil pour la diminution d'artéfacts de mouvement et du bruit lors du traitement d'images vidéo
EP1347415B1 (fr) Procédé d'accentuation d'image sans amplification du bruit
GB2337391A (en) Interlaced to progressive scanning conversion with edge enhancement by vertical temporal interpolation
EP1722330A2 (fr) Procédé et système d'extension des données d'images binaires pour les données d'images timbrées
US5225915A (en) Image processing with noise enhancing operators for moire reduction and/or random dot generation
US20030206662A1 (en) Method and apparatus for improving perceived digital image quality
JP2577748B2 (ja) 画像信号の補間方法及びそれを実施する画像信号処理装置
US4947260A (en) Method and apparatus for generating composite images
US6124841A (en) Image size enlarging and reducing method and apparatus
JP2555430B2 (ja) 画像の輪郭強調方法
JP4065462B2 (ja) 画像処理装置及び画像処理方法
JP3738791B2 (ja) 画像処理方法および装置
US6181834B1 (en) Hybrid image reduction method and apparatus with moir{acute over (e)} suppression
WO1996008930A1 (fr) Format yuv9 condense pour stockage imbrique et traitement efficace de donnees video numeriques
EP0814429A2 (fr) Système de traitement d'image
JP2002183727A (ja) 画像処理装置
JP2000209432A (ja) 画像処理方法
US6577778B1 (en) Method and apparatus for interpolating a digital image
US6104503A (en) Image processing and reproducing apparatus

Legal Events

Date Code Title Description
AK Designated states

Kind code of ref document: A1

Designated state(s): AM AT AU BB BG BR BY CA CH CN CZ DE DK EE ES FI GB GE HU IS JP KE KG KP KR KZ LK LR LT LU LV MD MG MN MW MX NO NZ PL PT RO RU SD SE SG SI SK TJ TM TT UA UG UZ VN

AL Designated countries for regional patents

Kind code of ref document: A1

Designated state(s): KE MW SD SZ UG AT BE CH DE DK ES FR GB GR IE IT LU MC NL PT SE BF BJ CF CG CI CM GA GN ML MR NE SN TD TG

DFPE Request for preliminary examination filed prior to expiration of 19th month from priority date (pct application filed before 20040101)
121 Ep: the epo has been informed by wipo that ep was designated in this application
WWE Wipo information: entry into national phase

Ref document number: 1995938826

Country of ref document: EP

WWP Wipo information: published in national office

Ref document number: 1995938826

Country of ref document: EP

REG Reference to national code

Ref country code: DE

Ref legal event code: 8642

WWW Wipo information: withdrawn in national office

Ref document number: 1995938826

Country of ref document: EP